flwrlabs/celeba

Dataset Card for Dataset Name

CelebFaces Attributes Dataset (CelebA) is a large-scale face attributes dataset with more than 200K celebrity images, each with 40 attribute annotations. The images in this dataset cover large pose variations and background clutter. CelebA has large diversities, large quantities, and rich annotations, including:

• 10,177 number of identities,

• 202,599 number of face images, and

• 5 landmark locations, 40 binary attributes annotations per image.

The dataset can be employed as the training and test sets for the following computer vision tasks: face attribute recognition, face recognition, face detection, landmark (or facial part) localization, and face editing & synthesis.

This dataset is used in Federated Learning research because of the possibility of dividing it according to the identities of the celebrities. This repository enables us to use it in this context due to the existence of celebrity id (celeb_id) beside the images and attributes.

Dataset Details

This dataset was created using the following data (all of which came from the original source of the dataset):

• aligned and cropped images (in PNG format),
• celebrities annotations,
• list attributes.

The dataset was divided according to the split specified by the authors (note the celebrities do not overlap between the splits).

Dataset Sources

• Website: https://liuziwei7.github.io/projects/FaceAttributes.html and https://mmlab.ie.cuhk.edu.hk/projects/CelebA.html
• Paper: Deep Learning Face Attributes in the Wild

Uses

In order to prepare the dataset for the FL settings, we recommend using Flower Dataset (flwr-datasets) for the dataset download and partitioning and Flower (flwr) for conducting FL experiments.

To partition the dataset, do the following.

1. Install the package.

pip install flwr-datasets[vision]

2. Use the HF Dataset under the hood in Flower Datasets.

from flwr_datasets import FederatedDataset
from flwr_datasets.partitioner import NaturalIdPartitioner

fds = FederatedDataset(
    dataset="flwrlabs/celeba",
    partitioners={"train": NaturalIdPartitioner(partition_by="celeb_id")}
)
partition = fds.load_partition(partition_id=0)

E.g., if you are following the LEAF paper, the target is the Smiling column.

Dataset Structure

Data Instances

The first instance of the train split is presented below:

{'image': <PIL.PngImagePlugin.PngImageFile image mode=RGB size=178x218>,
 'celeb_id': 1,
 '5_o_Clock_Shadow': True,
 'Arched_Eyebrows': False,
 'Attractive': False,
 'Bags_Under_Eyes': True,
 'Bald': False,
 'Bangs': False,
 'Big_Lips': False,
 'Big_Nose': False,
 'Black_Hair': False,
 'Blond_Hair': True,
 'Blurry': False,
 'Brown_Hair': True,
 'Bushy_Eyebrows': False,
 'Chubby': False,
 'Double_Chin': False,
 'Eyeglasses': False,
 'Goatee': False,
 'Gray_Hair': False,
 'Heavy_Makeup': False,
 'High_Cheekbones': True,
 'Male': True,
 'Mouth_Slightly_Open': True,
 'Mustache': False,
 'Narrow_Eyes': True,
 'No_Beard': True,
 'Oval_Face': False,
 'Pale_Skin': False,
 'Pointy_Nose': True,
 'Receding_Hairline': False,
 'Rosy_Cheeks': False,
 'Sideburns': False,
 'Smiling': True,
 'Straight_Hair': False,
 'Wavy_Hair': False,
 'Wearing_Earrings': False,
 'Wearing_Hat': False,
 'Wearing_Lipstick': False,
 'Wearing_Necklace': False,
 'Wearing_Necktie': False,
 'Young': False}

Data Splits

    train: Dataset({
        features: ['image', 'celeb_id', '5_o_Clock_Shadow', 'Arched_Eyebrows', 'Attractive', 'Bags_Under_Eyes', 'Bald', 'Bangs', 'Big_Lips', 'Big_Nose', 'Black_Hair', 'Blond_Hair', 'Blurry', 'Brown_Hair', 'Bushy_Eyebrows', 'Chubby', 'Double_Chin', 'Eyeglasses', 'Goatee', 'Gray_Hair', 'Heavy_Makeup', 'High_Cheekbones', 'Male', 'Mouth_Slightly_Open', 'Mustache', 'Narrow_Eyes', 'No_Beard', 'Oval_Face', 'Pale_Skin', 'Pointy_Nose', 'Receding_Hairline', 'Rosy_Cheeks', 'Sideburns', 'Smiling', 'Straight_Hair', 'Wavy_Hair', 'Wearing_Earrings', 'Wearing_Hat', 'Wearing_Lipstick', 'Wearing_Necklace', 'Wearing_Necktie', 'Young'],
        num_rows: 162770
    })
    valid: Dataset({
        features: ['image', 'celeb_id', '5_o_Clock_Shadow', 'Arched_Eyebrows', 'Attractive', 'Bags_Under_Eyes', 'Bald', 'Bangs', 'Big_Lips', 'Big_Nose', 'Black_Hair', 'Blond_Hair', 'Blurry', 'Brown_Hair', 'Bushy_Eyebrows', 'Chubby', 'Double_Chin', 'Eyeglasses', 'Goatee', 'Gray_Hair', 'Heavy_Makeup', 'High_Cheekbones', 'Male', 'Mouth_Slightly_Open', 'Mustache', 'Narrow_Eyes', 'No_Beard', 'Oval_Face', 'Pale_Skin', 'Pointy_Nose', 'Receding_Hairline', 'Rosy_Cheeks', 'Sideburns', 'Smiling', 'Straight_Hair', 'Wavy_Hair', 'Wearing_Earrings', 'Wearing_Hat', 'Wearing_Lipstick', 'Wearing_Necklace', 'Wearing_Necktie', 'Young'],
        num_rows: 19867
    })
    test: Dataset({
        features: ['image', 'celeb_id', '5_o_Clock_Shadow', 'Arched_Eyebrows', 'Attractive', 'Bags_Under_Eyes', 'Bald', 'Bangs', 'Big_Lips', 'Big_Nose', 'Black_Hair', 'Blond_Hair', 'Blurry', 'Brown_Hair', 'Bushy_Eyebrows', 'Chubby', 'Double_Chin', 'Eyeglasses', 'Goatee', 'Gray_Hair', 'Heavy_Makeup', 'High_Cheekbones', 'Male', 'Mouth_Slightly_Open', 'Mustache', 'Narrow_Eyes', 'No_Beard', 'Oval_Face', 'Pale_Skin', 'Pointy_Nose', 'Receding_Hairline', 'Rosy_Cheeks', 'Sideburns', 'Smiling', 'Straight_Hair', 'Wavy_Hair', 'Wearing_Earrings', 'Wearing_Hat', 'Wearing_Lipstick', 'Wearing_Necklace', 'Wearing_Necktie', 'Young'],
        num_rows: 19962
    })
})

Citation

When working with the CelebA dataset, please cite the original paper. If you're using this dataset with Flower Datasets and Flower, you can cite Flower.

BibTeX:

@inproceedings{liu2015faceattributes,
  title = {Deep Learning Face Attributes in the Wild},
  author = {Liu, Ziwei and Luo, Ping and Wang, Xiaogang and Tang, Xiaoou},
  booktitle = {Proceedings of International Conference on Computer Vision (ICCV)},
  month = {December},
  year = {2015} 
}

@article{DBLP:journals/corr/abs-2007-14390,
  author       = {Daniel J. Beutel and
                  Taner Topal and
                  Akhil Mathur and
                  Xinchi Qiu and
                  Titouan Parcollet and
                  Nicholas D. Lane},
  title        = {Flower: {A} Friendly Federated Learning Research Framework},
  journal      = {CoRR},
  volume       = {abs/2007.14390},
  year         = {2020},
  url          = {https://arxiv.org/abs/2007.14390},
  eprinttype    = {arXiv},
  eprint       = {2007.14390},
  timestamp    = {Mon, 03 Aug 2020 14:32:13 +0200},
  biburl       = {https://dblp.org/rec/journals/corr/abs-2007-14390.bib},
  bibsource    = {dblp computer science bibliography, https://dblp.org}
}

Dataset Card Contact

For questions about the dataset, please contact Ziwei Liu and Ping Luo. In case of any doubts about the dataset preparation, please contact Flower Labs.